Thursday, November 7, 2013

PSA: Janna Winrate

There are a lot of theories about why Janna's winrate is so good.  Some of them rely on random chance.  Some think Janna is strong.  Some rely on solo queue being a strange place where weird things happen.

Here's my take on Janna's winrate.

It all has to do with the role she plays and the algorithm that measures effectiveness in terms of skill.
(Nerdy math take at the bottom!!)

What do people play Janna as?  A support (except Zekent hue hue).

What is the least loved role?  Support.

What role is therefore played the most with the least emphasis and practice?  Support.

What I mean by this is that support is generally the least loved role and is therefore practiced the least by most league players.  This state of the game equates to a general lack of practice and therefore skill when it comes to the person playing the support role.  

So, support champions that have a low skill floor excel as supports overall in solo queue because there are a lot of people forced into playing support that don't practice the role.

It's the reason that Taric had such a high winrate before his nerfs and remake etc.  Taric has an incredibly low skill floor but his effectiveness was incredibly high.
And while Janna isn't nearly the same, the same idea applies to her.  Compared to other viable supports currently, the skill required to play Janna and get some effectiveness out of her is fairly low.

As a high elo support player once told me, being a good support is all about being less detrimental to your team than the other teams support.  And when it comes to solo queue, being less detrimental as Janna is simple in a lot of ways.
Janna is fast, her passive and her W let her move quickly.  This means you get caught less and feed kills less frequently, not to mention it's easy to re-position.
Her one skill shot is very wide and easy to hit when being pursued.  And while her ultimate can ruin games used incorrectly, as long as you sit next to your carry it's generally pretty easy to get some sort of good use from it.  Most of the time Janna ults gone wrong come from trying to make plays instead of just peel for your carry.

This is not to say that Janna has a low skill cap.  There are a ton of small things that separate good and bad Janna players like when to shield, using your ult to the optimal efficiency and maximizing tornado effectiveness.  But because Janna is a support these things matter much less towards her overall winrate because the nuances which make her harder to play aren't what the majority of supports using her have to worry about.  Because as long as you're shielding your carry and peeling for them to some extent (which is quite simple on Janna) you are doing quite a good job at your role already.  You might not be carrying your team but you are minimizing your detriment.  Essentially Janna is a good support to get carried as.

What I meant above about the algorithms:
f(x) = effectiveness and x = skill
Take a simple champion with a general lack of complexity but high return on low skill.  They might look something like this:
f(x) = x + 10
Now take a champion that is pretty much getting what you put in.  At lower levels you get less out but at higher levels you can get more.  That might look something like this:
f(x) = 5x
Now let's take a champion that without at least a good amount of skill isn't very effective but you get very high returns if you're skilled.
f(x) = x^2

So if we were to look at these three champions, the first would be most effective at a skill level of 1 or 2.  The second would be most effective at a skill level of 3 or 4.  The third would be on par with the second at 5 and superior at all other skill levels. 

No comments:

Post a Comment